"Opteron F SPEC Benchmarks"

[9-inch]

Awesome numbers for a 4-way Socket-F Opteron (can't quote it):

http://www.spec.org/osg/cpu2000/results/res2006q3/cpu2000-20060721-06586.html

 

[Action_Man]

Not as awesome as this

 

[battousai831]

posting spree i see today...

 

[ltcommander_data]

You know, I've been waiting for you to post this.

According to your post, the top of the line 4S/8P 2.8GHz 8220SE gets 146 on SPECint_rate_base2000. Fine.

However, a comparable top of the line 4S/8P Tulsa setup gets as follows:

http://www.spec.org/cpu2000/results/res2006q3/cpu2000-20060807-06940.html

162 in SPECint_rate_base2000.

Imagine that, your Opteron system is beaten by a Netburst processor. What's more, the Tulsa system was running a 32-bit OS while your Opteron was running a 64-bit OS. So the Opteron also had it's much publicized 64-bit advantage.

 

[Action_Man]

It'd be nice if we had a global ignore.

 

[battousai831]

that 1 made me lol...

 

[9-inch]

You know, I've been waiting for you to post this.

According to your post, the top of the line 4S/8P 2.8GHz 8220SE gets 146 on SPECint_rate_base2000. Fine.

However, a comparable top of the line 4S/8P Tulsa setup gets as follows:

http://www.spec.org/cpu2000/results/res2006q3/cpu2000-20060807-06940.html

162 in SPECint_rate_base2000.

Imagine that, your Opteron system is beaten by a Netburst processor. What's more, the Tulsa system was running a 32-bit OS while your Opteron was running a 64-bit OS. So the Opteron also had it's much publicized 64-bit advantage.

Uhmm, OK. That's fine.
Netburst were always good at int performance. Funny you didn't include your beloved woodcrest on that. Oops, sorry, Woodcrest can't go more than 4 sockets.

It's really lame that intel had to use a 16MB L3 cache to help out that darn hot Tulsa chip. :lol: :lol: :lol:

Socket-F will rape that processor at FP performance.

Nice try, but you didn't impress me.

 

[turpit]

Nice try, but you didn't impress me.

Just like you've failed to impress anyone in 687 posts under the 9 inch ID

 

[Action_Man]

Like your Turion X2 you're a failure sock puppet. BTW hows your mates BSer and sexbomb going?

 

[Action_Man]

How old are you 9nm?

 

[ltcommander_data]

It's really lame that intel had to use a 16MB L3 cache to help out that darn hot Tulsa chip.

16MB of shared L3 cache may be a ungraceful instrument, but since Intel has the extra die space of their 65nm process why not use it? Especially for a high end product where the costs are justified. In any case I would want to see some power numbers. You know, at 120W a pop, the those 2.8GHZ 8220SEs aren't power efficient either.

Socket-F will rape that processor at FP performance.

I haven't seem Tulsa SPECfp_rate_base2000 results yet, but I'll concede the point that K8 is stronger than Netburst at FP calculations. At least, Tulsa proves itself in Int calculations which means that customers can choose between the processors based on their needs.

 

[turpit]

How old are you 9nm?

1 year to match every nm?

 

[battousai831]

And another one locked.... http://forumz.tomshardware.com/hardware/quot-AMD-Aims-30-Global-Market-Share-quot-ftopict197341.html

Are you going for a record? This is now 36 total.

looks like it ><

 

[clairvoyant129]

That's nice. Want a cookie?

 

[1Tanker]

Jack, by golly....you've answered on of THGF's great mysteries. What does 9-Inch refer to? It's the length of his nose. KEWL.

 

[K8MAN]

Me waits for the 16 socket socket 1207 benchies. When QC hits that'll be 64 proc's/system. I'd love to see the numa memory scores of that beast with DDR2.

 

[johngoodwin]

You know, I've been waiting for you to post this.

According to your post, the top of the line 4S/8P 2.8GHz 8220SE gets 146 on SPECint_rate_base2000. Fine.

However, a comparable top of the line 4S/8P Tulsa setup gets as follows:

http://www.spec.org/cpu2000/results/res2006q3/cpu2000-20060807-06940.html

162 in SPECint_rate_base2000.

Imagine that, your Opteron system is beaten by a Netburst processor. What's more, the Tulsa system was running a 32-bit OS while your Opteron was running a 64-bit OS. So the Opteron also had it's much publicized 64-bit advantage.

I don't know if you noticed, but you also posted a benchmark where it was running 2x as many threads, and the xeon had 2x as much memory.

I'm not quite sure this was apples to apples.

Not to mention the OS is different.

John